Subscriber OnlyBooksReview

Bytes and Bullets: Study of AI arms race is meticulously researched and deeply disturbing

Steve Feldstein charts global technological-military rivalry whose implications are not fully understood

A 'Hunter Killer' drone at an expo in Kyiv. AI 'intelligentised warfare' and quantum changes in machine-led decision making have led to the escalation and lethality we have witnessed recently in conflicts from Ukraine to Gaza. Photograph: Brendan Hoffman/The New York Times
A 'Hunter Killer' drone at an expo in Kyiv. AI 'intelligentised warfare' and quantum changes in machine-led decision making have led to the escalation and lethality we have witnessed recently in conflicts from Ukraine to Gaza. Photograph: Brendan Hoffman/The New York Times
Bytes and Bullets: Global Rivalries, Private Tech and the Future of Modern Warfare
Author: Steve Feldstein
ISBN-13: 978-1800962866
Publisher: Monoray
Guideline Price: £25; 400pp

Steve Feldstein, a senior fellow at the Carnegie Endowment for International Peace, catalogues and captures in Bytes and Bullets what he describes as a critical moment of “tech transformation” in the pursuit of power and global dominance.

In the harnessing of artificial intelligence (AI) and advanced computing power, Feldstein describes a “fourth industrial revolution”, where world powers, particularly the United States and China, compete for geopolitical primacy through coercive algorithmic strategies and, worryingly, autonomous force projection.

The book charts the shift from a “unipolar” world of “stability” in the first quarter of the 21st century – what Feldstein calls the “post-cold war euphoria” of the 1990s – to the shock of the 9/11 attacks to the recent abandonment of the international laws and conventions of armed conflict by full-spectrum powers such as Russia and the United States.

In parallel with this dangerous trajectory, he describes an emerging global “military civil fusion” leading to AI “intelligentised warfare” with the quantum changes in machine-led decision making that have led to the escalation and lethality that we have witnessed recently in conflicts from Ukraine to Gaza.

Feldstein charts a rapid acceleration in the harnessing of Silicon Valley in what is referred to as the “military-tech gold rush” which began during US president Donald Trump’s first term in office from 2016 to 2020. Before Trump, the US had traditionally relied on the defence contractors Lockheed Martin, Boeing, Raytheon, Northrop Grumman and General Dynamics to underpin its military-industrial complex.

Over Trump’s two terms, the US government has aggressively recruited the tech industries – previously uninvolved in military applications – to collaborate with the Pentagon’s new Defence Innovation Unit, headed by the former senior Apple executive Doug Beek. In short, the Trump administration has fostered a “pro-military” Silicon Valley from what had previously been considered a “digital Switzerland” in terms of non-engagement with lethal applications.

To this end, tech companies such as Anduril, Palantir, Eprius, Hermeus, Shield AI and SpaceX have been recruited – along with international tech companies such as Germany’s Helsing – to create autonomous command and control systems and “precision” targeting and weapon systems that scrape world data centres to make decisions at a speed multiple times faster than human thought.

Feldstein lists the main players in the US, such as Palantir’s Peter Thiel – co-founder of the “Dialog group” network which had planned to meet in Powerscourt Hotel later this year – along with Anduril’s Palmer Luckey as being at the vanguard of America’s race for supremacy in AI machine-led “algorithmic warfare”.

The US has identified China as its primary competitor in this AI-enhanced digital arms race. Feldstein sets out how Xi Jinping has created China’s “Thousand Talents” programme with massive financial incentives and rewards for PhD and postdoctoral researchers to innovate in tech-enhanced warfare in ground, air, maritime and cyber domains.

Should decisions about whether AI can be used in war be left to private companies?Opens in new window ]

He also sets out how China’s industrial power gives it a distinct advantage in replicating, developing and scaling western advances in AI and computing power. He demonstrates the change in Beijing’s relationship with Moscow, with a catastrophic fall in status for Vladimir Putin from being a “partner” of Xi to being helplessly dependent on China’s military and technological aid for his criminal and failed invasion of Ukraine, and in turn becoming subordinated to China’s long-term geopolitical strategy in east and central Asia.

Feldstein makes the point that AI-enhanced, data-driven, algorithmic precision munitions and targeting systems carry with them the inherent risk of uncontrolled escalation – with swarms of cheaply manufactured lethal drones, controlled by “mega constellations” of GPS satellites, orchestrating the mass-killing of target populations, civilians and military alike, with no human involvement in the “kill-chain” process.

Feldstein observes that Trump’s adversarial and abrasive leadership style has alienated US allies and created levels of fear and uncertainty among its “axis of rivals” including China, Russia, Iran and North Korea that have led to a de facto arms race whose implications are not fully understood by scientific or military experts.

Key to these developments is the availability of hyperscale data-centre development. Countries such as Ireland, with – for now – plentiful water and cool temperate conditions, will be an essential component in the “data-scraping” aspect of these weapon systems. China has identified areas around Beijing, Shenzen and Shanghai as suitable for concentrating data centres.

Feldstein concludes this meticulously researched – and deeply disturbing – book by stating that AI and tech will certainly “enable” the human race to achieve outcomes in all areas of human activity – but whether these outcomes are liberating and positive or destructive is not for machines to decide. He argues that we have a narrowing window of opportunity to direct technology for good. For Ireland, there is an a priori requirement for us to decide our role – as a world-leading data and tech centre – in this AI arms race.

Tom Clonan is an Independent member of the Seanad and a former officer in the Irish Army